Global Optimization Algorithm And Its Application To SNP Association Analysis

Epistatic interactions between multiple single nucleotide polymorphisms (SNPs) arenow considered as a key factor to affect individual susceptibility to common diseases andmedicines, thus, the research on detection of epistatic interactions in SNP associationanalysis plays a pivotal role in disease diagnosis and cure as well as drug design. However,this is still a challenging and unsolved issue under the present situation of SNP data withhigh dimension and very limited number of samples.This paper is dedicated to research on the algorithm of detecting epistatic interactionin SNP association analysis. Through the research on the principle of two epistaticinteraction detection algorithms, i.e, AntEpiSeeker and SNPRuler, this paper provides anproposed AntEpiSeeker algorithm with the support of sub rule of locus sets as itsrelevance evaluation criteria, combined with the optimal solution search process ofAntEpiSeeker algorithm and the relevance evaluation criteria of SNPRuler algorithm.The effectiveness of this proposed algorithm is verified through massive experiments onsimulation data sets and real data sets. By the comparation and analysis of theexperiments of proposed algorithm with that of the AntEpiSeeker algorithm, it revealsthat the proposed algorithm has more advantages than AntEpiSeeker algorithm.